From the locust’s point of view

Variability across trials: is it necessary to average across neurons, or is there significant information in single neurons?
Is precise timing relationship between spikes in different neurons important in coding?